Adaptive Resource Management in the Cloud: The CORT (Cloud Open Resource Trading) Case Study

Essentially, cloud computing is based on a scenario where specialized providers offer a huge amount of resources to stakeholders, which require them based on their actual needs. The stakeholders needs may be characterized by significant fluctuations due to the model of payment, which is based on the pay-per-use paradigm. This has an important implication: cloud providers have to make an initial expensive investment in resources. Nowadays, available cloud computing solutions are provided by known players on the IT market such as Google, IBM, Amazon, or Microsoft. Small or medium size players have difficulties to compete in this domain. In this paper, we introduce our vision for cloud computing which aims to (1) exploit adaptive mechanisms for resource management and optimize resource usage by discouraging idle resources, (2) provide a market exchange model exploiting trading mechanisms for resource allocation, and (3) support failure management for critical situations. These objectives are sustained by a motivating example. Furthermore, the paper introduces several implementation hints based on our current prototype.